GAdaBoost: Accelerating Adaboost Feature Selection with Genetic Algorithms

TL;DR

GAdaboost introduces a genetic algorithm to efficiently select features for Adaboost, significantly reducing training time with minimal impact on detection accuracy in face detection tasks.

Contribution

It presents a novel method combining genetic algorithms with Adaboost to accelerate feature selection in object detection.

Findings

GAdaboost is up to 3.7 times faster than traditional Adaboost.

Detection accuracy decreases by only 3-4% on benchmark datasets.

The method maintains competitive detection performance with reduced training time.

Abstract

Boosted cascade of simple features, by Viola and Jones, is one of the most famous object detection frameworks. However, it suffers from a lengthy training process. This is due to the vast features space and the exhaustive search nature of Adaboost. In this paper we propose GAdaboost: a Genetic Algorithm to accelerate the training procedure through natural feature selection. Specifically, we propose to limit Adaboost search within a subset of the huge feature space, while evolving this subset following a Genetic Algorithm. Experiments demonstrate that our proposed GAdaboost is up to 3.7 times faster than Adaboost. We also demonstrate that the price of this speedup is a mere decrease (3%, 4%) in detection accuracy when tested on FDDB benchmark face detection set, and Caltech Web Faces respectively.